This is an investigation in statistical methodology of the robustness of four sequential procedures against departures from normality and independence: three sequential t-tests (open plan, restricted plan, wedge plan) and a non-parametric sequential signed-rank test. The situations under investigation were: uniform, scale contaminated normal, and Cauchy. These four test procedures were also examined under normal distributions with a specified correlation structure.